Deep Learning and Optimization Approaches in IoT-Based Smart Pharmacies for Optimizing Stock Management with Siamese Heterogeneous Convolutional Neural Networks: A Review

Abstract

The rapid evolution of digital healthcare systems has driven the integration of Internet of Things (IoT), artificial intelligence (AI), and deep learning technologies to enhance pharmaceutical supply chain management. Smart pharmacy systems utilize IoT-enabled devices, cloud platforms, and predictive analytics to improve inventory monitoring, reduce drug wastage, and ensure timely medication availability. Traditional inventory systems often rely on manual processes that are prone to errors, inaccurate demand forecasting, and inefficient stock utilization, leading to stockouts or overstocking. Recent advancements in deep learning, particularly convolutional neural networks (CNNs), have shown strong potential in analyzing complex healthcare data for inventory optimization. Siamese heterogeneous convolutional neural networks (SHCNNs) further enhance this capability by learning similarity relationships across diverse datasets such as pharmacy records, sensor data, and environmental inputs. IoT-based systems provide real-time visibility through sensors and RFID technologies, enabling automated decision-making for stock replenishment and expiry monitoring. The integration of AI with IoT improves operational efficiency and accuracy in inventory management. This review highlights key architectures and approaches, emphasizing the potential of hybrid intelligent systems while identifying challenges such as scalability, data security, and system integration for future research.